neurogastroenterology & motility - rome foundation...recommendations for pharmacological...

REVIEW ARTICLE

Recommendations for pharmacological clinical trials in

children with irritable bowel syndrome: the Rome

foundation pediatric subcommittee on clinical trials

M. SAPS,* M. A. L. VAN TILBURG,† J. V. LAVIGNE,‡,§,¶,** A. MIRANDA,†† M. A. BENNINGA,‡‡ J. A. TAMINIAU§§,# & C.

DI LORENZO*

*Division of Pediatric Gastroenterology, Hepatology & Nutrition, Nationwide Children’s Hospital, Columbus, OH, USA

†Division of Gastroenterology and Hepatology, Center for Functional Gastrointestinal and Motility Disorders, University of North

Carolina, Chapel Hill, NC, USA

‡Department of Child and Adolescent Psychiatry, Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL, USA

§Feinberg School of Medicine, Northwestern University, Chicago, IL, USA

¶Mary Ann and J. Milburn Smith Child Health Research Program, Chicago, IL, USA

**Children’s Hospital of Chicago Research Center, Chicago, IL, USA

††Division of Pediatric Gastroenterology Hepatology & Nutrition, Medical College of Wisconsin, Milwaukee, WI, USA

‡‡Department of Pediatric Gastroenterology and Nutrition, Emma Children’s Hospital/Academic Medical Center, Amsterdam,

The Netherlands

§§Member of the Pediatric Committee (PDCO) European Medicines Agency, London, UK

Key Points

• The Rome Foundation subcommittee for Pharmacological Clinical Trials in Children with IBS recommends

randomized, double-blind, placebo-controlled, parallel-group, clinical trials to assess the efficacy of new drugs.

• Entry criteria for abdominal pain are aweekly average ofworst abdominal pain in past 24 h of at least 3.0 on a 0–10point scale or at least 30 mm in 100 mmVisual Analog Scale and for stool endpoints an average stool consistency

of lower than 3 in the Bristol Stool Form Scale (BSFS) during the run-in period for clinical trials on IBS-C and an

average stool consistency of greater than 5 in the BSFS during the run-in period for clinical trials on IBS-D.

• The endpoints for abdominal pain are a decrease in intensity of at least 30% compared with baseline and to

meet or exceed the Reliable Change Index (RCI) for the sample.

• Changes in stool consistency are the primary endpoints for both, IBS with diarrhea (IBS-D) and IBS with

constipation (IBS-C) in children.

Abstract

Background There is little published evidence of

efficacy for the most commonly used treatments.

Thus, there is an urgent need to conduct clinical trials

on existing and novel therapies. Purpose In order to

address these issues the Rome Foundation and mem-

bers of the Pediatric Committee of the European

Medicines Agency formed a subcommittee on clinical

trials to develop guidelines for the design of clinical

Address for Correspondence

Miguel Saps, MD, Department of Pediatrics, Division ofGastroenterology, Hepatology & Nutrition, NationwideChildren’s Hospital, 700 Children’s Drive, Columbus, OH43205, USA.Tel: +1 (614) 722-3457; fax: + (614) 722-3454;e-mail: [email protected]#The opinion is the personal view of the author and does notrepresent the agency’s position.Received: 10 February 2016Accepted for publication: 30 May 2016

© 2016 John Wiley & Sons Ltd 1

Neurogastroenterol Motil (2016) doi: 10.1111/nmo.12896

Neurogastroenterology & Motility

trials in children with irritable bowel syndrome (IBS).

The following recommendations are based on evi-

dence from published data when available and expert

opinion. Key recommendations The subcommittee

recommends randomized, double-blind, placebo-con-

trolled, parallel-group, clinical trials to assess the

efficacy of new drugs. The combined endpoints for

abdominal pain are a decrease in intensity of at least

30% compared with baseline and to meet or exceed

the Reliable Change Index (RCI) for the sample. Stool

consistency is measured with the Bristol Stool Scale

Form (BSFS). The subcommittee recommends as entry

criteria for abdominal pain a weekly average of worst

abdominal pain in past 24 h of at least 3.0 on a 0–10point scale or at least 30 mm in 100 mm Visual

Analog Scale. For stool endpoints the committee

recommends an average stool consistency lower than

3 in the BSFS during the run-in period for clinical trials

on IBS-C and an average stool consistency greater

than 5 in the BSFS during the run-in period for clinical

trials on IBS-D. Changes in stool consistency are the

primary endpoints for both IBS with diarrhea (IBS-D)

and IBS with constipation (IBS-C).

Keywords abdominal pain, children, clinical trials,

endpoints, irritable bowel syndrome, stool consis-

tency.

INTRODUCTION

Chronic abdominal pain is one of the symptoms most

commonly encountered by pediatric providers and

gastroenterologists. Abdominal pain or discomfort is

the hallmark of functional abdominal pain disorders

(FAPDs), conditions which include irritable bowel

syndrome (IBS), functional dyspepsia (FD), functional

abdominal pain – not otherwise specified (FAP-NOS)

and abdominal migraine. The development of pharma-

cological therapies for these disorders in both adults

and children has been limited by several factors, such

as the incomplete understanding of their peripheral

and central pathophysiological mechanisms, the lack

of actionable biomarkers, the heterogeneity of these

disorders and the major discrepancies in the method-

ologies and endpoints used in clinical trials. Thus, it is

not surprising that here are currently no approved

drugs for the treatment of FAPDs in children and there

is little published evidence of efficacy for the most

commonly used treatments. Only a few, small, ran-

domized clinical trials have evaluated the effect of

pharmacological interventions in the treatment of

FAPDs in children1 and thus, there is an urgent need

to conduct clinical trials on existing and novel thera-

pies.

Most drug clinical trials in adult patients with IBS2–15

and some of the largest trials on FAPDs in children16

have used global outcomes measures. In April 2009, the

US Foods and Drug Administration (FDA) proposed the

use of provisional primary endpoints for clinical trials in

adults with IBS in place of global outcome measures. A

guideline recommending the use of co-primary end-

points, assessing two major aspects of IBS – abnormal

defecation and abdominal pain – was published by the

FDA in 2012.17 The use of these co-primary endpoints

was subsequently also adopted by the European

Medicines Agency (EMA) in 2013.18 It remains unclear

whether these recommendations are adequate for

studies in children.19 Specific considerations in the

pediatric population frequently preclude using the same

methodology in adults and children. Developmental

and cognitive limitations and differences in clinical

relevance between adults and childrenmake some adult

endpoints poorly applicable to children. Different

symptom scales are validated in children of different

age. Parents are also often hesitant to enroll their

children in lengthy randomized trials that include a

placebo arm.

In order to address these issues, the Rome Founda-

tion and members of the Pediatric Committee of the

European Medicines Agency formed a subcommittee

on clinical trials that was charged with developing

guidelines for the design of pediatric clinical trials in

children with IBS. This subcommittee conducted a

comprehensive review of the English scientific litera-

ture on clinical trials in functional gastrointestinal

disorders (FGIDs) and pediatric chronic pain scales and

stool scales. In addition, the subcommittee sought to

identify gaps in knowledge about pediatric pain and

stool scales to outline the future research agenda. The

following recommendations are based on evidence

from published data when available and expert opin-

ion. The subcommittee considered that there was

insufficient data to provide recommendations that

could be based exclusively on high level of evidence.

Therefore, the recommendations issued by this sub-

committee could potentially be modified in future

editions as new data become available.

RECOMMENDATION

Recommendations for Study Design

The goal of a clinical trial is to assess the safety and

efficacy of an intervention to relieve or decrease the

severity of the child’s symptoms, to reduce the impact

© 2016 John Wiley & Sons Ltd2

M. Saps et al. Neurogastroenterology and Motility

of the disorders on the child’s life and to improve the

quality of life of the patient. Existing drug trials in

FAPDs in children have used a wide variety of study

designs, length of treatment, definitions, outcomes and

measures. Disparities among studies make it difficult

to compare results and combine data for meta-analysis.

Randomized, double-blind, placebo-controlled, par-

allel-group, clinical trials are recommended to assess

the efficacy of new drugs. Pharmacological therapy for

FAPDs in children has been associated with high

placebo response rates that can range from a nocebo

effect to more than 50% in some studies.20 Therefore,

double-blind, randomized, placebo-controlled, parallel

group clinical trials are recommended to assess the

efficacy of new drugs. Parallel design in which each

study participant is randomly assigned to a group

(intervention or placebo) with all the participants in

each group receiving or not receiving the intervention

is recommended. Parallel design avoids possible ‘order’

effects in which the effect or side effects of a drug could

affect the outcome and possible ‘carry-over’ effects of

crossover trials. Although the ‘carry-over’ effects could

be minimized by a prolonged ‘wash-out’ period

between treatments, this would prolong the length of

the trial and result in lower recruitment and higher

attrition rates. These challenges outweigh the benefits

of crossover design which include avoiding possible

imbalances between groups and needing a lower sam-

ple size. A common challenge in conducting random-

ized controlled trials (RCT) in children is successful

recruitment of large number of patients. In order to

overcome this challenge, multicenter trials should be

considered.

A period of baseline assessment without treatment

(‘run-in period’) is recommended. It is recommended

that all trials start with a period of baseline assessment

without treatment (‘run-in period’). A run-in period of

at least 1 week (preferably 2 weeks) is suggested to

help screen out ineligible participants and provide

objective baseline information.

Study duration should be of at least 4 weeks (prefer-

ably 6 weeks or more). The duration of the trial to best

assess efficacy of a treatment was a topic of much

debate within the working group. The duration should

be sufficient to not only adequately address the

outcome being measured, but also potential side effects

or adverse reactions that could result from the therapy.

It should also not be so long that children and their

families become overburdened. Longer trial periods

may also limit recruitment, a very common problem in

pediatric clinical trials. It is recommended that the

study duration should be at least 4 weeks (preferably

6 weeks or more). Duration of the trial should be based

on the pharmacology of the compound. As a general

rule, treatment periods shorter than 4 weeks are not

recommended due to the variable course of IBS with

periods of ‘waxes and wanes’. A monitored treatment

free period at the end of the trial is recommended to

assess whether beneficial effects are maintained once

the intervention is discontinued.

Recommendations for Patient Selection

Selection of subjects should best reflect the population

that is affected. Demographic variables of patients

included or excluded from the trial should be docu-

mented and the criteria for restricting the study

population must be justified. Recruitment of subjects

with broad demographic features is advisable. This

includes different age, ethnicity, race, and gender.

Enrolling subjects with varied degrees of pain inten-

sity/duration and bowel symptoms is important in

order to have a wide representation of the disorder in

the community and because of common non-treat-

ment effects in RCT such as regression to the mean,

ceiling and bottom effects. Enrolling children who

meet Rome criteria but have mild average pain is

discouraged (e.g., <3.0 on a 10-point numerical pain

intensity scale or <30 mm in 100 mm Visual Analog

Scale [VAS]) since such a low baseline score may not

be sufficient to demonstrate clinically meaningful

improvement.21

It is best to identify different forms of FAPDs, such

as IBS and FAP-NOS (functional dyspepsia is not

covered in this guidelines), and perform trials only on

those patients. As signs and symptoms of constipa-

tion predominant IBS (IBS-C) and diarrhea predomi-

nant IBS (IBS-D) differ, each type of disorder should

be studied in separate clinical trials. The pediatric

Rome IV committees have defined for the first time

diagnostic criteria for IBS-C and IBS-D (Table 1). It is

recommended to include specific IBS sub-types, such

that IBS-C and IBS-D patients are only studied with

pharmacological treatments appropriate to their sub-

type.

The general recommendation is to enroll children

8 years of age or older for the purpose of clinical trials,

as most self-report measures are approved at this age.

As there is little agreement between parent and child

report of symptoms22 the committee recommends that

child report should be used for children ages 8 years old

and above or up to the lowest age for which the pain

scale has been validated. When enrolling children that

are younger than 8 years of age, proxy report (parent or

caretaker) may be required based on the validated

questionnaire that is used.


Rome foundation recommendations

Prior to the entry into the trial, all participants

should receive education, information and reassurance

as standard care for IBS. With the exception of

necessary aspects inherent to the protocol such as

contacting patient, follow up visits, documentation

and possible workup, normal practice of care should

not be modified and patients should be managed

following normal clinical practice.

A careful history should be obtained to select

patients that fulfill the criteria established by the

Rome IV criteria for IBS. The Rome criteria have been

widely accepted and used in RCT of children with

FGIDs. While there is evidence of the construct validity

of the Rome III criteria,23,24 there is also conflicting

evidence on the reliability of these criteria.25,26 The

Rome III criteria were found to be sensitive but their

specificity has been put into question.27 No validation

studies have been yet published on the new version of

the criteria (Rome IV), although studies are currently

being conducted. In addition to fulfilling the Rome

criteria, the inclusion criteria should also include a

focused history and diagnostic work-up. Alarm symp-

toms should be identified to rule out other disorders that

can mimic IBS.27–29

Required laboratory, imaging and endoscopic testing

should be specified prior to the initiation of the trial. At

a minimum, the work-up in children who present with

abdominal pain should include blood count, CRP, anti-

tissue transglutaminase antibody, IgA and determina-

tion of occult blood in stool and fecal calprotectin and/

or lactoferrin. Children with diarrhea should also

undergo stool culture and determination of ova and

parasites. Abnormal laboratory tests or the presence of

alarm features should prompt further investigation

even when meeting Rome criteria for IBS. In addition,

it is advisable to document any known history of co-

morbid psychological disorders and the presence of

additional gastrointestinal symptoms, as these factors

may influence treatment outcomes.

Similar principles should be applied to children with

a diagnosis of ‘functional abdominal pain-not other-

wise specified’ (NOS) (FAP-NOS) according to the

Rome IV criteria. This group of children should use

similar inclusion/exclusion criteria, trial design and

workup with the exception of the aspects that are not

pertinent to the diagnosis (changes in stool character-

istics).

The recommended inclusion and exclusion criteria

are summarized below.

Inclusion criteria

1 Satisfy Rome IV criteria (IBS-C, IBS-D or FAP-NOS).

2 Patient ages 8–18 years (lower ages can be considered

if appropriate measures for are available).

3 Patient and parent ability to read and comprehend

questionnaires.

4 Average daily pain rate of at least 3 out of 10

(numerical rating scale, NRS) or at least 30 mm in a

100 mm VAS met during a run-in period of at least

1 week (or similar cut off if a different validated pain

scale is used for the study).

Exclusion criteria

1 Children who tested positive for bacterial or para-

sites infections.

2 Carbohydrate malabsorption, diagnosed either clini-

cally (2 weeks exclusion diet with resolution of

symptoms) or with proper testing (breath test,30).

Children with carbohydrate intolerance who con-

tinue to have IBS symptoms while on an exclusion

diet can still be included.

3 Children with chronic gastrointestinal disorders that

mimic FAPDs: inflammatory bowel disease, pancre-

atitis, chronic liver disease, eosinophilic esophagitis,

peptic ulcer disease, celiac disease, pseudo-obstruc-

tion, small bowel bacterial overgrowth, or Hirsch-

sprung’s disease.

4 Significant chronic health condition requiring spe-

cialty care (e.g., lithiasis, ureteropelvic junction

obstruction, sickle cell, cerebral palsy, hepatic,

hematopoietic, renal, endocrine, or metabolic dis-

eases) that could potentially impact the child’s ability

to participate or confound the results of the study.

5 Unintentional weight loss greater than or equal to

5% of their body weight within the last 3 months.

6 Decreased growth velocity.

7 Gastrointestinal blood loss.

8 Recurrent or unexplained fevers.

9 Pregnancy.

10 Developmental disabilities impairing ability to

understand or communicate.

11 History of hypersensitivity or allergy to medication

being tested.

12 History of previous abdominal surgeries in the past

3 months.

13 Rome IV criteria diagnosis of functional constipa-

tion.

Table 1 Rome criteria IV – irritable bowel syndrome (IBS) – IBS-C and

IBS-D Subtypes

IBS with predominant constipation (IBS-C): >¼ (25%) of bowel

movements with Bristol stool types 1 or 2 and <¼ (25%) bowel

movements with Bristol stool types 6 or 7

IBS with predominant diarrhea (IBS-D): >¼ (25%) of bowel movements

with Bristol stool types 6 or 7 and <¼ (25%) bowel movements with

Bristol stool types 1 or 2

IBS, irritable bowel syndrome. Based on the Bristol Stool Form Scale

(BSFS).21



Recommendations for Concurrent Use ofMedications or Therapies

The use of concurrent medications that affect pain

sensitivity or psychiatric disorders should be carefully

evaluated. Patients using drugs that affect intestinal

function or pain sensation should generally be

excluded. Patients on other medications at the time

of screening should be encouraged to maintain a

constant dose and schedule during the entire trial if

medically possible. Concurrent medication(s) should

be taken during the run in period in order to assess the

effect on the baseline measures. Consideration should

be given to the mechanism of action of the drug being

studied and interference with metabolism of other

drugs that may mask or potentiate the effect of drug

being studied. Patients undergoing specific diets can be

included provided no changes in diet are expected

within the study period. Subjects should not start any

new medications, complementary or alternative ther-

apies during the study period. If the patient’s status

requires a new intervention or treatment, researchers

should evaluate the impact of such changes and

consider stopping participation in the current trial if

ethically appropriate.

Recommendations for Documentation

Demographic information on patients entered and

excluded (gender, age, race, ethnicity, site of enroll-

ment), and reasons for exclusion, should be docu-

mented. We recommend the inclusion of children with

a single functional gastrointestinal diagnosis. How-

ever, if children with co-existence of two diagnoses are

included in the study the presence of both should be

documented as the outcome of this subset of patients

may differ from those with a single diagnosis. Simi-

larly, we recommend documenting the presence of

extra-intestinal somatic complaints.

When available, psychological disturbances should

be evaluated and the information documented. Psy-

chological comorbidities could influence the outcome

resulting in more or less favorable treatment

response.31,32

The use of daily diaries is recommended. Daily

diaries should include relevant study data preferably in

electronic form. The use of daily diaries, when possible

with reminder alarms, helps minimize recall bias25and

records the date and time of completion of the

questionnaire.

All adverse events should be documented and

reported as unexpected adverse eventsmay occur during

the course of the trial. Stopping rules have to be pre-

established and documented in the study protocol. All

adverse events should be evaluated by an independent

data and safety monitoring board (DSMB). For general

guidelines see: https://www.nichd.nih.gov/health/clin-

icalresearch/clinical-researchers/steps/Pages/conduct_

monitor.aspx; http://www.fda.gov/OHRMS/DOCKETS/

98fr/01d-0489-gdl0003.pdf.

Treatment allocation and randomization should be

specified and documented a priori including method of

random allocation sequence and type of randomiza-

tion. Prior to initiation, trials should be registered in a

public location (i.e. https://clinicaltrials.gov/; https://

www.clinicaltrialsregister.eu/). The results of the

study should be published even if the results of the

trial are negative or inconclusive. Sources of funding

and conflicts of interest of each investigator should be

disclosed.

Recommendations for Primary Endpoints

Primary endpoints should be based on patient reported

outcomes (PROs) when possible. Primary endpoints

should be based on patient reported outcomes (PROs)

in children 8 years of age and older and on proxy

reported outcomes (parents or other caretakers) in

younger children. Endpoints of clinical trials have

traditionally measured changes in pain and bowel

symptoms as well as measures of impact of these

symptoms on the child’s life such as quality of life and

disability. However, there is no agreement on what

constitutes an acceptable level of change in these

symptoms,33 nor there is agreement on which of the

several different approaches to measuring change (an-

chor-based, distribution-based, or others) should be

used.34–36

Recommendations for PRO and Proxy ReportedOutcome for Abdominal Pain

Many different approaches have been suggested for

PROs, including a global measure of change with

treatment, a meaningful clinical important difference

(MCID), and a percentage change in symptoms21 As

discussed previously, the EMA18 and the FDA17 rec-

ommended against the use of global outcome measures

on the change in IBS symptoms such as adequate relief

of symptoms with treatment (Table 2). In agreement

with the regulatory agencies, this committee consid-

ered that the exclusive use of a global endpoints lacked

specificity.

An alternative approach is to determine a MCID

based on interviews with children, parents or physi-

cians to determine a reduction in pain that corresponds



https://www.nichd.nih.gov/health/clinicalresearch/clinical-researchers/steps/Pages/conduct_monitor.aspx



http://www.fda.gov/OHRMS/DOCKETS/98fr/01d-0489-gdl0003.pdf

http://www.fda.gov/OHRMS/DOCKETS/98fr/01d-0489-gdl0003.pdf

https://clinicaltrials.gov/

https://www.clinicaltrialsregister.eu/

https://www.clinicaltrialsregister.eu/

with satisfactory pain control. Several studies have

yielded MCIDs for pediatric pain measures37but most

have been for acute pain and none have been conducted

in pediatric IBS. Despite the strength of the reliance on

child report of meaningful change, ratings of ‘better’

may vary with age33and this approach does not take

into account that an absolute change may have a

different meaning for children with different severity

of symptoms. Hence, the committee recommends

using a percentage change in pain instead.

The EMA and FDA have adopted the use of an

improvement in abdominal pain equal or greater than

30% as primary endpoint for IBS in adult clinical

trials.17,18 It remains unclear if the endpoints recom-

mended for adult studies are suitable for use in children.

No trials have examined this. The subcommittee con-

sidered the EMA suggestion that a higher percentage of

improvement of abdominal pain would be a preferred

endpoint in children. No published evidence was found

to substantiate this recommendation. A study compar-

ing the use of at least 30% and at least 50% improve-

ment in abdominal pain in children38 showed that the

use of the criterion: at least 50% improvementwasmore

specific in detecting a positive response to a pain relief

global question than the at least 30% criterion but the

sensitivity of the at least 50% improvement in abdom-

inal pain criterion was low (40%). Thus, the use of at

least 50% improvement in abdominal pain as primary

efficacy endpoint in clinical trials would result in a large

proportion of children that considered their symptoms

adequately relieved having negative results in clinical

trials.

No PRO measure is perfectly reliable across admin-

istrations, and the test–retest reliability of a measure

may vary as the interval between administrations

increases.34 Furthermore, due to developmental

changes across childhood, children of different ages

may vary considerably in their ability to report pain

reliably over time. In order to measure change reliably

using PROS, the reliability of the measure needs to be

considered. A recent study has shown that less than

half of subjects characterized as responders using the at

least 30% criterion achieved a reliable change based on

test–retest reliability and standard deviation of the pain

measure.33 Accounting for the test–retest reliability

and standard deviation of pain measures can be done by

calculating the Reliable Change Index (RCI; see

appendix for formula) which is a measure of change

in standardized units and it indicates the strength as

well as the direction of the change. A change in pain

that exceeds the RCI means the child is significantly

improved. Thus, the committee recommends that the

endpoints for change should be a decrease of abdominal

pain intensity of at least 30% from baseline and that

this value should be at least equal to the RCI for that

sample. Examples of this process are provided in the

appendix.

Applying these criteria require knowledge of the

test–retest reliability of the pain measure to be used

and standard deviation of the sample, before the start of

the trial. However, the committee understands that

there may be times when such estimates are not

available and a standard estimate of the RCI is needed.

Thus, the subcommittee decided to provide an

Table 2 Primary endpoints, entry and responder definition criteria for clinical trials on pediatric IBS*

Indication Primary endpoints Entry criteria Responder definition

IBS-C Abdominal pain intensity

AND stool consistency

Abdominal pain intensity

Weekly average of worst abdominal pain in

past 24 h ≥3.0 on a 0–10 point scale or ≥30 mm

in 100 mm Visual Analog Scale†

AND

Stool consistency

Bowel movements during run-in period with

average consistency <3 on the Bristol Stool

Form Scale: Type 1 (very hard) or Type 2 (hard)

Abdominal pain intensity (Dual-criteria)

≥30% improvement in abdominal pain

AND

Improvement ≥Reliable Change Index (RCI) at the

last week of trial compared with baseline

AND

Stool consistency

Improvement in (≥1 Bristol Stool Form Scale to a

higher number) average consistency at the last

week of trial compared with baseline

IBS-D Abdominal pain intensity

AND stool consistency

Abdominal pain intensity

Weekly average of worst abdominal pain in

past 24 h ≥3.0 on a 0–10 point scale or ≥30 mm

in 100 mm Visual Analog Scale†

AND

Stool consistency

Bowel movements during run-in period with

average consistency >5 on the Bristol Stool

Form Scale: Type 6 (loose), 7 (very loose).

Abdominal pain intensity (Dual-criteria)

≥ 30% improvement in abdominal pain

AND

Improvement ≥Reliable Change Index (RCI) at the

last week of trial compared with baseline

AND

Stool consistency

Improvement in (≥1 Bristol Stool Form Scale to a

lower number) average consistency at the last

week of trial compared with baseline

*Abdominal pain endpoints alone to be used for trials on FAP-NOS. †Or similar cut off for moderate pain on different pain scale if used for PRO in the

study.



alternative endpoint to be used in cases that the

calculation of the test–retest reliability of the pain

measure and the standard deviation of the sample is

not logistically possible. Based on the only study

determining RCI among children ages 8–17 years with

IBS, the RCI was 21.23 mm using a 0–100 mm VAS for

abdominal pain37 and exceeded the ≥30%change rule. In

order to allow for variation among samples, the Rome

Foundation Pediatric Subcommittee on Clinical Trials

believes that a RCI of at least 25 mm should be used as

abdominal pain endpoint. In order to assure enough pain

severity to participate in a trial, the committee recom-

mends a daily average minimum abdominal pain inten-

sity of at least 30 mm on a 100 VAS or at least 3.0 out of

10 on a numeric rating scale (0–10).Thus, for pain measures, the committee recom-

mends the PRO to be based on a dual standard: change

in pain meeting or exceeding both, 30% change in

intensity from baseline and the RCI for that sample.

An acceptable alternative to this approach, especially

when no data is available to determine the RCI before

the trial, is to use a cut-off of 25 mm change on a VAS.

Additional Recommendations Related toMeasurement of Pain

1 Validated pain scales should be used for the assess-

ment of chronic pain in children. These include the

VAS and NRS or other scales with appropriate

validation.39–41

2 As children vary widely in their ability to recall

pain,42 the committee recommends collecting daily

24-h recall of pain for at least 7 days. Pain intensity

should be averaged over these 7 days. Daily diaries

should be collected at least 7 days before the start of

the trial and 7 days at the end of the trial while the

child is still on medication. Minimum amount of

time for trials is discussed elsewhere in this docu-

ment.

Recommendations for PRO and Proxy ReportedOutcome for Stool Consistency

Changes in stool consistency are the primary end-

points for both, IBS-D and IBS-C in children. The

EMA18 and FDA17 proposed two co-primary endpoints

for adults with IBS, the assessment of pain and stool

characteristics. Stool assessment recommendations

vary by IBS subtype. Both agencies proposed the

assessment of stool consistency, measured by Bristol

Stool Form Scale (BSFS) as outcome measure for IBS-D

and stool frequency for IBS-C.17,18 The committee

considered that consistency and not frequency was a

more relevant outcome in children with IBS-C and IBS-

D. Thus, the committee recommends changes in stool

consistency as primary endpoint for both, IBS-D and

IBS-C in children. The trial may consider the use of

frequency instead of consistency as primary endpoint if

the predominant or exclusive action of the drug is to

alter the frequency of bowel movements. The com-

mittee recommends as entry criteria an average stool

consistency lower than 3 in the BSFS during the run-in

period for clinical trials on IBS-C and an average stool

consistency greater than 5 in the BSFS during the run-

in period for clinical trials on IBS-D. Whenever possi-

ble and in every case in children younger than 8 years

of age, parents or caretakers should rate stool consis-

tency instead of the children. Recording of stool

characteristics should be done as close to the bowel

movement as possible to avoid recall bias.

Responder is defined as a patient who during the last

week of the trial achieved an average change in stool

consistency from baseline of at least one form in the

BSFS (to a harder stool consistency in IBS-D and to a

softer stool consistency in IBS-C). The committee

recognizes that the validation of the BSFS in children is

insufficient and the lack of data on MCID in stools in

children. The use of the modified BSFS43 was consid-

ered but the committee did not find enough evidence

to recommend this tool or evidence that was superior

to the original BSFS. The use of stool scales with a

lower number of categories should be considered with

caution as it may result in children requiring a greater

change in stool characteristics to demonstrate benefit.

This would result in studies that require a larger

sample size to demonstrate a significant difference

what may constitute an important limitation in pedi-

atric clinical trials that frequently struggle to enroll a

large number of children.

Recommendations for Secondary Measures

Although pain and stool consistency are the primary

measures for assessing treatment outcomes, the sub-

committee believes that the primary measure should

ideally be complemented by secondary measures. The

most important secondary measure for clinical trials in

IBS is disability. A large part of the treatment of IBS is

to reduce disability in children. Pediatric studies

conducted on children with FADPs have shown that

pain improvement does not always correlate with

improvements of disability.19 Children are frequently

encouraged to attend school and participate in age

appropriate activities no matter the level of pain.

Therefore, it is highly recommended to include a

validated measure of disability in children.



Other secondary measures of interest include pain

frequency, equal or greater than 50% improvement in

pain intensity, bowel movement frequency, no longer

meeting Rome criteria for the condition being studied,

quality of life and sleep. In addition, it is highly

recommended to include in each trial parental report of

primary and secondary outcome measures. Stool fre-

quency and straining with bowel movements may be

considered as secondary endpoints in children with

IBS-C and IBS-D. Some pharmacological interventions

may require the assessment of specific outcomes or

mechanistic factors depending on the characteristics of

the drug.

Recommendations for Data Analysis

Upon completion of the trial, the analysis should

assess the proportion of patients in each treatment arm

who fulfill a pre-established treatment responder def-

inition that represents a clinically meaningful change

to the patient (Table 3).

It is recommended that the analysis be conducted

using an intention-to-treat principle, in which data

from all patients enrolled are analyzed based on initial

treatment assignment regardless of their completion of

the trial or compliance with the protocol. The use of

per-protocol analysis could be valuable as secondary

analysis.

Interim analysis is not recommended. Interim anal-

ysis may jeopardize the integrity of the clinical trial and

result in reporting of inaccurate observations. Interim

analysis is justifiedwhen it is believed that participation

in the trial may expose participants at risk.

Sample size calculation is required and assumptions

for its calculation should be specified. Power calcula-

tions should be based on differences in proportions.

The calculation of the sample size should be clinically

relevant (powered to detect the MCID) and based on

the expected behavior of the primary endpoints (ex-

pected difference in proportions between groups).

Information on expected minimum effect size in

primary outcomes between groups, type I error level,

statistical power, and standard deviation of the differ-

ence if continuous outcomes will be considered should

be provided. Lack of sample size calculation or lack of

information on how the sample size is calculated is a

common problem in pediatric RCTs in IBS.1 An

insufficient sample size may explain the negative

results of some of the studies. At the time of sample

size calculation, it should be considered that a high

placebo effect may be present. Placebo effect in chil-

dren across RCTs has been variable with one trial

Table 3 Summary of recommendations

General recommendations� Primary endpoints should be based on patient reported outcomes (PROs) when possible.� Entry criteria should be based on the Rome IV criteria.� The use of daily diaries is recommended.� Results should be evaluated using intention-to-treat principle.� Interim analysis should be avoided.� Selection of subjects should reflect the population that is affected.� Demographic information on patients entered and excluded and reasons for exclusion should be documented.� Demographic and clinical characteristics of the subjects in the medication and placebo group should be documented. Laboratory,

imaging and endoscopic testing should be specified prior to the initiation of the trial.� All adverse events should be documented and reported and be evaluated by a DSMB.� Stopping rules have to be pre-established and documented in the study protocol.� Treatment allocation and randomization should be specified and documented a priori.� Trials should be registered in a public location.� The results of the study should be published even if the results of the trial are negative or inconclusive.� Sources of funding and conflicts of interest of each investigator should be disclosed.� Participants should receive education, information and reassurance as standard care for IBS.� Recording of stool characteristics should be done as close to the bowel movement as possible.� Dropouts timing and reasons should be documented.� Sample size calculation should be done and assumptions should be specified.� For reporting of results the CONSORT guidelines should be followed.Study design� Randomized, double-blind (or triple-blind), placebo-controlled, parallel-group, clinical trial.� Run-in period of at least 1 week.� Study duration of at least 4 weeks.� Two primary endpoints: abdominal pain intensity and changes in stool consistency (IBS-D and IBS-C) to be assessed as average

value at the last week of the trial.

Patients with constipation predominant IBS (IBS-C) and diarrhea predominant IBS (IBS-D) should be studied in separate clinical trials. Similar

recommendations apply to RCTs in children with a diagnosis of functional abdominal pain-not otherwise specified’ (NOS) (FAP-NOS) with the

exception of changes in stool characteristics.



showing a nocebo effect,20 while other studies had a

beneficial placebo ranging from 36% to 53%.1 The

multicenter trial by Saps et al. reported a placebo effect

of 53% but a placebo effect as high as 75% was found if

any improvement in the non-intervention arm would

have been considered as a positive effect. RCTs for IBS

in children had an attrition rate up to 19.4%.44

Adjustments for dropout rate and loss to follow-up

should be done. The procedures used for adjustment

should be noted. Dropouts timing and when possible

reasons should be documented.

CONSORT GUIDELINES

Studies should provide a detailed flow diagram and

follow the CONSORT guidelines (http://www.consort-

statement.org/).

COMMENTS

The current document is the culmination of the work

of the Rome Foundation and members of the Pediatric

Committee of the European Medicines Agency who

aimed to bring standardization to the field and suggest

best practices for pharmacological trials in children

and adolescents.

Despite the recent emphasis on PROs, there is

surprisingly little evidence in the pediatric literature

to guide the design of clinical trials and the definition

of clinically relevant and reliable outcome measures in

pediatric IBS. There is considerable variation in entry

criteria, length of duration of trial, and outcome

measures in current pediatric trials.1 Design and

outcome measures for each trial either follow recom-

mendations for adults with IBS or borrow heavily from

acute pain trials. Neither approach is optimal as

decisions for adults may not generalize well to children

and pediatric acute pain measures have not been

validated for IBS.

In a pediatric population one should be particularly

concerned about reliability of measures in younger

children in whom cognitive abilities to recall and report

symptoms are still developing.25 In addition, children

may lack reading skills needed to complete question-

naires. Studies have shown good validity and reliability

of several measures of acute pain, including non-verbal

scales.40,45,46 These scales have widely been used in the

pediatric chronic pain literature. However, we do not

know the reliability of these scales in chronic pain

patients. For IBS patients, pain waxes and47 and there-

fore needs to be assessed over several days to weeks,

which raises issues of optimal timing for recall, duration

of data collection to ensure representative data andwhat

constitutes a meaningful difference in pain over time.

Unfortunately, the subcommittee found no data on the

validity and reliability of pain measures over prolonged

periods of time and on what may constitute a meaning-

ful difference in reduction of pain for pediatric IBS

patients or chronic pediatric pain in general. We have

even less data on stool endpoints for IBS. Similarly to the

pain endpoint, there are few published data, and most

scales (such as the BSFS) are not well validated in

children48 nor are meaningful differences defined. This

gap in the literature hampered the subcommittee’s

ability to make evidence based recommendations on

any of the current scales.

One study has tried to fill this void in the literature

and determined meaningful differences of abdominal

pain in a large study of pediatric IBS patients and found

validation for defining a responder as at least 30%

reduction in pain combined with exceeding the RCI of

the pains33for a VAS this was defined as at least 25 mm

reduction. This study is encouraging as it provides an

empirical established estimate of a clinically meaning-

ful difference. However, this study is in need of

replication before we can recommend these cut-offs

with any degree of certainty. Furthermore, this study

used only one pain measure (a VAS of pain), and other

measures may be of interest to researchers depending

on the study population. For example, researchers may

need to include non-verbal measure of pain in younger

children, and estimates of meaningful clinical differ-

ences in chronic pain are currently not available for

these measures. Based on the lack of evidence in

pediatric pain measures, the subcommittee highly

recommends for researchers embarking on trial design

to collect additional data on reliability (RCI) of pain

measures in children and adolescents with IBS.

In light of the lack of evidence on reliability of pain

measures in children and adolescents with IBS the

subcommittee recommends a two-pronged approach.

The subcommittee recommends using the >30%change plus exceeding RCI rule. In order to do so,

investigators need to collect additional data on relia-

bility of their pain measures before embarking on

trials. This not only ensures that trial data is reliable

and meaningful, but also that more data will be

collected on pediatric pain measures which will inform

future trial design. If collection of reliability data is not

feasible, the subcommittee recommends investigators

use cut-offs determined in a previous trial of IBS using

the VAS (≥25 mm change represents a responder).

Our recommendations on stool endpoints differ from

EMA and FDA guidelines that recommend stool

consistency in IBS-D and stool frequency in IBS-C as

endpoints. The subcommittee considered that



http://www.consort-statement.org/

http://www.consort-statement.org/

consistency is a more meaningful outcome for children

and that stool consistency was more clearly associated

with the Rome definition of IBS-C and IBS-D that

requires having hard stools or loose stools in more than

25% of days respectively. In addition, the subcommit-

tee has polled authors of studies on pediatric IBS-C to

obtain information on the number of daily bowel

movements in children with this diagnosis. Data from

two studies have shown that the mean number of

bowel movements was variable among children with

IBS-C with a wide spectrum that ranged from one to

seven BMs a week.48,49 These results are in agreement

with a published study on IBS-subtypes in children that

showed that the majority of children with IBS-C had >3bowel movements weekly with 48% of children having

one or more bowel movements daily.50 These studies

demonstrate that in children with IBS-C the frequency

of bowel movements is highly variable and there is an

overlap between the number of bowel movements in

children with IBS-C and healthy children. Moreover,

the use of number of bowel movements as endpoint in

children that may be attending school during the trial

and could withhold their stools either because they are

not allowed to leave the classroom or to avoid the

embarrassment of being noticed by their peers leaving

the classroom to go to the bathroom could affect the

number of bowel movements and provide inaccurate

information. Children who have multiple bowel move-

ments s a day may also not be able to accurately recall

the number of bowel movements at the end of the day.

In addition to the primary PRO pain and stool, the

subcommittee also considered important to measure

disability. Reductions in disability (e.g., school avoid-

ance) are important outcomes in IBS trials and may be

independent of reductions in symptoms. Other sec-

ondary measures to consider were discussed. The

subcommittee also included recommendations for trial

design, largely based on best practices adjusted to the

needs and ability of children. These include recom-

mendations for study inclusion/exclusion criteria,

measurement periods, child vs parental reports of

PRO, trial duration, randomization, blinding etc.

FUTURE RECOMMENDATIONS

It is obvious that there are large gaps in knowledge in

this field. Clearly, we are in need of studies to guide

future recommendations. One of the most immediate

pressing issues is to examine validity and stability

over longer periods of time for the most common pain

and stool measures. These studies should use inter-

vals similar to those in RCTs. This information is

needed in order to be able to assess the expected

fluctuations in IBS symptoms that occur naturally and

to estimate the strength of treatment effects when

taking those fluctuations into account. Furthermore,

studies are needed to determine developmentally

appropriate meaningful differences in these measures,

parent-child concordance of symptoms and how to

optimize daily diaries of symptoms (e.g., by electronic

diary). Studies on the effect of comorbid symptoms

and the effect of somatization should be conducted to

better understand the influence of these factors on the

outcomes of clinical trials and to characterize subsets

of patients that may achieve a greater benefit from

each intervention. The effect of placebo in children

and the factors that may influence its effect are in

need of further investigation. We call on pediatric

investigators to make trial design and measurement a

focus of their study.

FUNDING

No funding declared.

DISCLOSURES

Miguel Saps: Scientific consultant Forest, Quintiles, Ardelyx,QOL Medical, IMHealth Science, Nutricia. Miranda A.L. vanTilburg: Takeda (research support). John Lavigne: No disclosures.Miranda Adrian: Scientific consultant QOL Medical, Marc Ben-ninga: Scientific consultant Shire, Sucampo, Astrazeneca, Nor-gine, Zeria, Danone, Friesland Campina, Novalac. Jan Taminiau:No disclosures. Carlo Di Lorenzo: Scientific consultant QOLMedical, IMHealth Science, Merck.

AUTHOR CONTRIBUTIONS

Miguel Saps designed the study, contributed to the analysis ofthe published data, contributed to the development of recom-mended guidelines, wrote the paper, and critically reviewed themanuscript. Miranda A. L. van Tilburg designed the study,contributed to the analysis of the published data, contributed tothe development of recommended guidelines, wrote the paper,and critically reviewed the manuscript. John Lavigne designedthe study, contributed to the analysis of the published data,contributed to the development of recommended guidelines,wrote the paper, and critically reviewed the manuscript. AdrianMiranda designed the study, contributed to the analysis of thepublished data, contributed to the development of recommendedguidelines, wrote the paper, and critically reviewed the manu-script. Marc Benninga designed the study, contributed to theanalysis of the published data, contributed to the development ofrecommended guidelines, wrote the paper, and criticallyreviewed the manuscript. Jan Taminiau critically reviewed theliterature and the manuscript. Carlo Di Lorenzo designed thestudy, contributed to the analysis of the published data,contributed to the development of recommended guidelines,wrote the paper and critically reviewed the manuscript.



REFERENCES

1 Saps M, Biring HS, Pusatcioglu CK,Mintjens S, Rzeznikiewiz D. A com-prehensive review of randomized pla-cebo-controlled pharmacologicalclinical trials in children with func-tional abdominal pain disorders. J

Pediatr Gastroenterol Nutr 2015; 60:645–53.

2 Camilleri M, Chey WY, Mayer EA,Northcutt AR, Heath A, Dukes GE,McSorley D, Mangel AM. A random-ized controlled clinical trial of theserotonin type 3 receptor antagonistalosetron in women with diarrhea-predominant irritable bowel syn-drome. Arch Intern Med 2001; 161:1733–40.

3 Camilleri M, Mayer EA, DrossmanDA, Heath A, Dukes GE, McSorley D,Kong S, Mangel AW et al. Improve-ment in pain and bowel function infemale irritable bowel patients withalosetron, a 5-HT. Aliment Pharma-

col Ther 1999; 13: 1149–59.4 Camilleri M, Northcutt AR, Kong S,

Dukes GE, McSorley D, Mangel AW.Efficacy and safety of alosetron inwomenwith irritable bowel syndrome:a randomised, placebo-controlled trial.Lancet 2000; 355: 1035–40.

5 Chang L, Ameen VZ, Dukes GE,McSorley DJ, Carter EG, Mayer EA.A dose-ranging, phase II study of theefficacy and safety of alosetron in menwith diarrhea-predominant IBS. Am JGastroenterol 2005; 100: 115–23.

6 Chey WD, Chey WY, Heath AT,Dukes GE, Carter EG, Northcutt A,Ameen VZ. Long-term safety andefficacy of alosetron in women withsevere diarrhea-predominant irritablebowel syndrome. Am J Gastroenterol2004; 99: 2195–203.

7 Fried M, Beglinger C, Bobalj NG,Minor N, Coello N, Michetti P;TegaSwiss Study Group. Tegaserod issafe, well tolerated and effective inthe treatment of patients with non-diarrhoea irritable bowel syndrome.Eur J Gastro Hepatol 2005; 17: 421–7.

8 Jones RH, Holtmann G, Rodrigo L,Ehsanullah RS, Crompton PM, Jac-ques LA, Mills JG. Alosetron relievespain and improves bowel functioncompared with mebeverine in femalenonconstipated irritable bowel syn-drome patients. Aliment PharmacolTher 1999; 13: 1419–28.

9 Kellow J, Lee OY, Chang FY, Thong-sawat S, Mazlam MZ, Yuen H, Gwee

KA, Bak YT et al. An Asia-Pacific,double blind, placebo controlled, ran-domised study to evaluate the efficacy,safety, and tolerability of tegaserod inpatients with irritable bowel syn-drome. Gut 2003; 52: 671–6.

10 M€uller-Lissner SA, Fumagalli I, Bard-han KD, Pace F, Pecher E, Nault B,R€uegg P. Tegaserod, a 5-HT4 receptorpartial agonist, relieves symptoms inirritable bowel syndrome patientswith abdominal pain, bloating andconstipation. Aliment PharmacolTher 2001; 15: 1655–66.

11 M€uller-Lissner S, Holtmann G,Rueegg P, Weidinger G, L€offler H.Tegaserod is effective in the initialand retreatment of irritable bowelsyndrome with constipation. Aliment

Pharmacol Ther 2005; 21: 11–20.12 M€uller-Lissner S, Koch G, Talley NJ,

Drossman D, Rueegg P, Dunger-Bal-dauf C, Lefkowitz M. Subject’s Glo-bal Assessment of Relief: anappropriate method to assess theimpact of treatment on irritablebowel syndrome-related symptomsin clinical trials. J Clin Epidemiol2003; 56: 310–6.

13 Novick J, Miner P, Krause R, GlebasK, Bliesath H, Ligozio G, R€uegg P,Lefkowitz M. A randomized, double-blind, placebo-controlled trial of tega-serod in female patients sufferingfrom irritable bowel syndrome withconstipation. Aliment PharmacolTher 2002; 16: 1877–88.

14 Nyhlin H, Bang C, Elsborg L, Silven-noinen J, Holme I, R€uegg P, Jones J,Wagner A. A double-blind, placebo-controlled, randomized study to eval-uate the efficacy, safety and tolerabil-ity of tegaserod in patients withirritable bowel syndrome. Scand J

Gastroenterol 2004; 39: 119–26.15 Tack J, M€uller-Lissner S, Bytzer P,

Corinaldesi R, Chang L, Viegas A,Schnekenbuehl S, Dunger-Baldauf Cet al. A randomised controlled trialassessing the efficacy and safety ofrepeated tegaserod therapy in womenwith irritable bowel syndrome withconstipation. Gut 2005; 54: 1707–13.

16 Saps M, Youssef N, Miranda A,Nurko S, Hyman P, Cocjin J, DiLorenzo C. Multicenter, randomized,placebo-controlled trial of amitripty-line in children with functionalgastrointestinal disorders. Gastro-

enterology 2009; 137: 1261–9.17 U.S. Department of Health and

Human Services Food and Drug

Administration Center for Drug Eva-luation and Research (CDER). Gui-dance for industry: irritable bowelsyndrome – clinical evaluation ofdrugs for treatment. 2012. Availableat: http://www.fda.gov/downloads/Drugs/Guidances/UCM205269.pdf

18 Committee for Medicinal Productsfor Human Use. Guideline on theevaluation of medicinal products forthe treatment of irritable bowel syn-drome. CPMP/EWP/785/97 2014;Rev. 1: 1–18. Available at: http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guide-line/2013/07/WC500146176.pdf

19 Mohammad S, Di Lorenzo C, YoussefNN, Miranda A, Nurko S, Hyman P,Saps M. Assessment of abdominalpain through global outcomes andrecent FDA recommendations in chil-dren. Are we ready for change?J Pediatr Gastroenterol Nutr 2014;58: 46.

20 Bahar RJ, Collins BS, Steinmetz B,Ament ME. Double-blind placebo-controlled trial of amitriptyline forthe treatment of irritable bowel syn-drome in adolescents. J Pediatr 2008;152: 685–9.

21 Lavigne J, Saps M. When is painimprovement clinically meaningful?– First study on Minimum ClinicallyImportant Difference (MCID) in chil-dren with Abdominal Pain (AP_ –Predominant Functional Gastrointes-tianl Disorders (AP-FGIDS). Gas-

troenterology 2015; 148: S-587.22 Garber J, Van Slyke DA, Walker LS.

Concordance between mothers’ andchildren’s reports of somatic andemotional symptoms in patients withrecurrent abdominal pain or emo-tional disorders. J Abnorm Child

Psychol 1998; 26: 381–91.23 Caplan A, Walker L, Rasquin A.

Validation of the pediatric Rome IIcriteria for functional gastrointestinaldisorders using the questionnaire onpediatric gastrointestinal symptoms.J Pediatr Gastroenterol Nutr 2005;41: 305–16.

24 Saps M, Nichols-Vinueza DX, Mint-jens S, Pusatcioglu CK, Velasco-Ben�ıtez CA. Construct validity of thepediatric Rome III criteria. J Pediatr

Gastroenterol Nutr 2014; 59: 577–81.25 Chogle A, Sztainberg M, Bass L,

Youssef NN, Miranda A, Nurko S,Hyman P, Cocjin J et al. Accuracy ofpain recall in children. J Pediatr Gas-troenterol Nutr 2012; 55: 288–91.



http://www.fda.gov/downloads/Drugs/Guidances/UCM205269.pdf

http://www.fda.gov/downloads/Drugs/Guidances/UCM205269.pdf

http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2013/07/WC500146176.pdf




26 Van Tilburg M, Squires M, Blois-Mar-tinN,LeibyA,LangsederA.Test of thechild/adolescent Rome III criteria:agreement with physician diagnosisand daily symptoms. Neurogastroen-terol Motil 2013; 25: 302–e246.

27 Gijsbers CF, Benninga MA, Sch-weizer JJ, Kneepkens CF, VergouweY, B€uller HA. Validation of the RomeIII criteria and alarm symptoms forrecurrent abdominal pain in children.J Pediatr Gastroenterol Nutr 2014;58: 779–85.

28 El-Chammas K, Majeskie A, SimpsonP, Sood M, Miranda A. Red flags inchildren with chronic abdominal painand crohn’s disease – a single centerexperience. J Pediatr 2013; 162: 783–7.

29 El-Matary W, Spray C, Sandhu B.Irritable bowel syndrome: the com-monest cause of recurrent abdominalpain in children. Eur J Pediatr 2004;163: 584–8.

30 Gremse DA, Greer AS, Vacik J,Dipalma JA. Abdominal pain associ-ated with lactose ingestion in chil-dren with lactose intolerance. Clin

Pediatr 2003; 42: 341–5.31 Kaczynski KJ, Simons LE, Claar RL.

Anxiety, coping, and disability: a testof mediation in a pediatric chronicpain sample. J Pediatr Psychol 2011;36: 932–41.

32 Lavigne JV, Saps M, Bryant FB. Mod-els of anxiety, depression, somatiza-tion, and coping as predictors ofabdominal pain in a community sam-ple of school-age children. J Pediatr

Psychol 2013; 39: 9–22.33 Saps M, Lavigne J. Abdominal pain

endpoints currently recommended bythe FDA and EMA for adult patientswith irritable bowel syndrome maynot be reliable in children. Neurogas-troenterol Motil 2015; 27: 849–55.

34 American Education Research Asso-ciation APANCoMiE. The Standards

for Educational and PsychologicalTesting. Washington, DC: American

Education Research AssociationAPANCoMiE, 2014.

35 Cella D, Eton DT, Lai J-S, PetermanAH, Merkel DE. Combining anchorand distribution-based methods toderive minimal clinically importantdifferences on the Functional Assess-ment of Cancer Therapy (FACT) ane-mia and fatigue scales. J Pain

Symptom Manage 2002; 24: 547–61.36 Lavigne JV. Systematic review: issues

in measuring clinically meaningfulchange in self-reported chronic pedi-atric pain intensity. J Pediatr Psychol2016; pii: jsv161. [Epub ahead of print]

37 Tsze DS, Hirschfeld G, Baeyer CL,Bulloch B, Dayan PS. Clinically sig-nificant differences in acute painmeasured on self-report pain scalesin children. Acad Emerg Med 2015;22: 415–22.

38 Mohammad S, Pusatcioglu C, SapsM.Comparison of primary efficacy end-points recommended by regulatoryagencies in children with functionalgastrointestinal disorders. Gastroen-

terology 2015; 148: S-586.39 Bailey B, Bergeron S, Gravel J, Daoust

R. Comparison of four pain scales inchildren with acute abdominal painin a pediatric emergency department.Ann Emerg Med 2007; 50: 379–383.e372.

40 Bailey B, Daoust R, Doyon-Trottier E,Dauphin-Pierre S, Gravel J. Valida-tion and properties of the verbalnumeric scale in children with acutepain. PAIN� 2010; 149: 216–21.

41 von BaeyerCL, Spagrud LJ,McCormickJC, Choo E, Neville K, Connelly MA.Three new datasets supporting use ofthe Numerical Rating Scale (NRS-11)for children’s self-reports of pain inten-sity. PAIN� 2009; 143: 223–7.

42 Self M, Williams A, Czyzewski D,Weidler E, Shulman R. Agreementbetween prospective diary data andretrospective questionnaire report ofabdominal pain and stooling symp-toms in children with irritable bowel

syndrome. Neurogastroenterol Motil2015; 27: 1110–9.

43 Chumpitazi BP, LaneMM,CzyzewskiDI, Weidler EM, Swank PR, ShulmanRJ. Creation and initial evaluation of aStool FormScale for children. J Pediatr2010; 157: 594–7.

44 Sadeghian M, Farahmand F, FallahiG, Abbasi A. Cyproheptadine forthe treatment of functional abdom-inal pain in childhood: a double-blinded randomized placebo-con-trolled trial. Minerva Pediatr 2008;60: 1367–74.

45 Hicks CL, von Baeyer CL, SpaffordPA, van Korlaar I, Goodenough B. TheFaces Pain Scale–Revised: toward acommon metric in pediatric painmeasurement. Pain 2001; 93: 173–83.

46 Pag�e MG, Katz J, Stinson J, Isaac L,Martin-Pichora AL, Campbell F. Val-idation of the numerical rating scalefor pain intensity and unpleasantnessin pediatric acute postoperative pain:sensitivity to change over time. J Pain2012; 13: 359–69.

47 van Lisman-Leeuwen Y, Spee LA,Benninga MA, Bierma-Zeinstra SM,Berger MY. Prognosis of abdominalpain in children in primary care—aprospective cohort study. Ann Fam

Med 2013; 11: 238–44.48 Giannetti E, de’Angelis G, Turco R,

Campanozzi A, Pensabene L, Salva-tore S, de Seta F, Staiano A. Subtypesof irritable bowel syndrome in chil-dren: prevalence at diagnosis and atfollow-up. J Pediatr 2014; 164: 1099–103. e1091.

49 Self MM, Czyzewski DI, ChumpitaziBP, Weidler EM, Shulman RJ. Sub-types of irritable bowel syndrome inchildren and adolescents. Clin Gas-

troenterol Hepatol. 2014; 12: 1468–73.

50 Rajindrajith S, Devanarayana NM.Subtypes and symptomatology of irri-table bowel syndrome in children andadolescents: a school-based surveyusing Rome III criteria. 2012.



APPENDIX

CALCULATING THE PRO BASED UPONTHE DUAL CRITERIA OF EXCEEDING THE30% IMPROVEMENT AND RCI VALUE

Procedures for calculating the RCI are described by

Jacobsen and Truax (1992). Table 1 illustrates RCIs for

various combination of test–retest correlations and

standard deviations of the pre-treatment sample that

will apply to most clinical trials. To illustrate, in a

Table A1 (A) RCI of the Visual Analog Scale. (B) RCI for the Faces Pain

Scale

SD value

Test–retest correlation

0.7 0.75 0.77 0.8 0.85 0.9 0.95

(A)

10 15.18 13.85 13.29 12.4 10.73 8.76 6.2

15 22.77 20.79 19.94 18.59 16.1 13.14 9.29

20 30.64 27.71 26.59 24.79 21.47 17.53 12.4

30 45.54 41.58 39.88 37.18 32.2 26.3 18.59

(B)

1 1.51 1.38 1.23 1.24 1.07 0.88 0.62

1.5 2.28 2.08 1.99 1.86 1.61 1.31 0.93

2 3.03 2.77 2.69 2.4 2.15 1.75 1.24

Table A2 Dual criteria PRO for VAS

VAS score

(baseline)

FDA

30%

30% reduction

value RCI*

PRO based on

dual criteria

100 0.3 30 25 30

90 0.3 27 25 27

80 0.3 24 25 25

70 0.3 21 25 25

60 0.3 18 25 25

50 0.3 15 25 25

40 0.3 12 25 25

30 0.3 9 25 25

20 0.3 6 25 25

10 0.3 3 25 25

*Estimated for a trial in which the test–retest reliability of the VAS is

.80 and the sample SD is 20. The 24.79 RCI value (Table 1) was

rounded to 25.

Table A3 Dual criteria PRO for Faces Pain Scale

FACES/NRS

(initial)

FDA

30%

30%

reduction

value RCI*

MCID based on

dual criteria

Rounded

MCID

10 0.3 3 1.86 3 3

9 0.3 2.7 1.86 2.7 3

8 0.3 2.4 1.86 2.4 3

7 0.3 2.1 1.86 2.4 3

6 0.3 1.8 1.86 1.86 2

5 0.3 1.5 1.86 1.86 2

4 0.3 1.2 1.86 1.86 2

3 0.3 0.9 1.86 1.86 2

2 0.3 0.6 1.86 1.86 2

1 0.3 0.3 1.86 NA NA

*Estimated for a trial in which the test–retest reliability of the scale is

.80 and the sample SD is 1.5. in the last column, the 1.86 RCI value

was rounded to 2.

sample in which the test–retest reliability is .8 and the

SD for the pretreatment sample is 20, the RCI would be

24.79. As scales vary widely in their scoring, the

standard deviation may vary, as illustrated for the VAS

(figure 1a) and Faces Pain Scale (A1 [B]).

Table 2 illustrates the process of combining the RCI

and the 30% FDA criteria to achieve the dual standard

PRO recommended by the subcommittee. As illus-

trated, the magnitude of the final value reported by the

child to meet the 30% change standard varies with the

pretreatment (baseline) score. In Table A2, the RCI

was based on a sample in which the test reliability is

0.80 and the pre-treatment SD is 20. In that instance, as

Table A2 shows, the dual-criteria PRO for the child

whose initial VAS score was 100 would be 30 (because

the 30% change score is greater than the RCI). For a

child with an initial VAS score of 50, the MCID would

be 25 (because the RCI is greater than the 30% change

criteria). Thus, a child with an initial VAS of 50 would

need to change by ≥25 to be classified as a responder.

Again, a similar process can be used for other mea-

sures, such as the revised Faces scale (See Table A3).



neurogastroenterology & motility - rome foundation...recommendations for pharmacological...

Documents